Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

DrAttack

Benchmarks

Task NameDataset NameSOTA ResultTrend
Jailbreak DetectionDrAttack
Accuracy99
30
Jailbreak DefenseDrAttack
ASR0
22
Safety EvaluationDrAttack
Safety Score95.49
16
Jailbreak DetectionDrAttack
ASR3
15
Showing 4 of 4 rows